Vol. 87: 57–66, 2009 DISEASES OF AQUATIC ORGANISMS Published November 16 doi: 10.3354/dao02084 Dis Aquat Org

Contribution to DAO Special 5 ’The role of environment and microorganisms in diseases of corals’ OPENPEN ACCESSCCESS Catabolite regulation of enzymatic activities in a white pox pathogen and commensal during growth on mucus polymers from the coral palmata

Cory J. Krediet1, Kim B. Ritchie2, 3, Max Teplitski1, 3,*

1Interdisciplinary Ecology Graduate Program, School of Natural Resources and Environment, University of Florida, Institute of Food and Agricultural Sciences (IFAS), 103 Black Hall, PO Box 116455, Gainesville, Florida 32611, USA 2Mote Marine Laboratory, 1600 Ken Thompson Parkway, Sarasota, Florida 34236, USA 3Department of Soil and Water Science, University of Florida, IFAS, 106 Newell Hall, PO Box 110510, Gainesville, Florida 32611, USA

ABSTRACT: Colonization of host mucus surfaces is one of the first steps in the establishment of coral- associated microbial communities. Coral mucus contains a sulfated glycoprotein (in which oligosac- charide decorations are connected to the polypeptide backbone by a mannose residue) and mole- cules that result from its degradation. Mucus is utilized as a growth substrate by commensal and pathogenic organisms. Two representative coral commensals, Photobacterium mandapamensis and meridiana, differed from a white pox pathogen Serratia marcescens PDL100 in the pat- tern with which they utilized mucus polymers of Acropora palmata. Incubation with the mucus poly- mer increased mannopyranosidase activity in S. marcescens, suggestive of its ability to cleave off oligosaccharide side chains. With the exception of glucosidase and N-acetyl galactosaminidase, gly- cosidases in S. marcescens were subject to catabolite regulation by galactose, glucose, arabinose, mannose and N-acetyl-glucosamine. In commensal P. mandapamensis, at least 10 glycosidases were modestly induced during incubation on coral mucus. Galactose, arabinose, mannose, but not glucose or N-acetyl-glucosamine had a repressive effect on glycosidases in P. mandapamensis. Incubation with the mucus polymers upregulated 3 enzymatic activities in H. meridiana; glucose and galactose appear to be the preferred carbon source in this bacterium. Although all these bacteria were capable of producing the same glycosidases, the differences in the preferred carbon sources and patterns of enzymatic activities induced during growth on the mucus polymer in the presence of these carbon sources suggest that to establish themselves within the coral mucus surface layer commensals and pathogens rely on different enzymatic activities.

KEY WORDS: Coral microbiology · Mucus surface polysaccharide · Coral pathogen

Resale or republication not permitted without written consent of the publisher

INTRODUCTION degrade components of the host surface-associated mucus (Vine et al. 2004, Fabich et al. 2008). In studies Attachment, colonization and establishment on host of bacteria associated with fish and corals, pathogens mucus surfaces are the first steps in the interactions rapidly degraded host surface mucus and reached high between bacteria and their eukaryotic hosts (Kurz et numbers (Vine et al. 2004, Sharon & Rosenberg 2008). al. 2003, Ritchie & Smith 2004, Rosenberg & Falkovitz When grown on host mucus in vitro, populations of 2004, Nehme et al. 2007). Pathogens and commensals commensal bacteria stabilized at densities that were at differ in the way they colonize host surfaces and least an order of magnitude lower than those reached

*Corresponding author. Email: [email protected] © Inter-Research 2009 · www.int-res.com 58 Dis Aquat Org 87: 57–66, 2009

by pathogens (Vine et al. 2004). Similarly, Vibrio spp., lipids make up 4.2% of the mucus (Ducklow & Mitchell which are related to coral pathogens, dominated 1979, Meikle et al. 1988). The chemical composition of microbial communities formed on mucus of the coral the mucus glycoprotein differs among coral species Oculina patagonica after an extended incubation in (Ducklow & Mitchell 1979, Molchanova et al. 1985, vitro (Rosenberg & Falkovitz 2004, Sharon & Rosen- Meikle et al. 1987, 1988, Klaus et al. 2007). Mucus of berg 2008), even though vibrios make up less than 5% Acropora formosa, for example, contains 36 to 38% of culturable bacteria in the mucus layer of this coral neutral sugars, 18 to 22% amino sugars and 19 to 30% under normal conditions (Koren & Rosenberg 2006). amino acids (Meikle et al. 1988). The oligosaccharide Collectively, these observations support the hypothesis decorations (2 to 4 sugar residues in length) are that in the absence of other factors, opportunistic attached to the polypeptide backbone by an O-glyco- pathogens can overgrow and dominate coral-associ- sidic linkage to serine or threonine through the carbon ated microbial communities, thus contributing to the 1 of mannose (Meikle et al. 1987). The glycoproteins appearance of disease signs (Ritchie 2006). from A. formosa contain terminal arabinose residues This hypothesis also implies that opportunistic linked by a β 1→3 bond (Meikle et al. 1987). The entire pathogens of corals have the ability to outcompete spectrum of enzymatic activities involved in the degra- native coral-associated bacteria under the conditions dation of coral mucus by pathogenic, commensal and that favor this overgrowth. Consistent with this hypo- environmental bacteria is not yet characterized. thesis, of all the tested coral and environmental bacte- Preparations of coral mucus contain a significant ria, Serratia marcescens PDL100 reached the highest amount of low molecular weight compounds (Meikle et final population density when grown in vitro on mucus al. 1987, 1988, Krediet et al. 2009). While the chemical of Acropora palmata (Krediet et al. 2009). This strain of nature of these compounds is not yet known, the low S. marcescens was originally isolated as a cause of molecular weight fraction of mucus most likely contains white pox, a rapidly progressing coral tissue necrosis simple sugars and oligosaccharides that resulted from of the threatened Caribbean coral A. palmata (Patter- enzymatic degradation of the polymer. Even though son et al. 2002). The observations that the white pox these compounds appear to be plentiful in mucus, their strain was more efficient at utilizing mucus of the host effect on enzymatic activities in coral mucus-associated coral in vitro suggested that this ability to efficiently bacteria is not yet known. Kuntz et al. (2005) hinted at a degrade mucus may be linked to its pathogenicity potential role of sugars in coral ecology: exposure of (Krediet et al. 2009). coral fragments to sugars at 5 to 25 mg l–1 led to the ap- The mechanisms by which Serratia marcescens pearance of various disease signs. For example, a 30 d PDL100 colonizes, infects, and causes disease in corals treatment with lactose (at 25 mg l–1), a disaccharide that are not yet clear. The progression of infections caused by is not known to occur in the coral reef environment, led Serratia spp. in other invertebrates suggests that to nearly 100% mortality of the Montastraea annularis colonization of host surfaces is the first in a series of fragments; however, survivorship of Porites furcata was events leading to the appearance of disease signs. For only modestly reduced by the treatment (Kuntz et al. example, to cause disease in nematodes and flies, S. 2005). A 30 d incubation with mannose caused signifi- marcescens first colonizes the intestines, degrades cells cant mortality in P. furcata but not in M. annularis of the alimentary tract, and then spreads to other organs (Kuntz et al. 2005). It is not yet clear whether these ob- (Kurz et al. 2003, Nehme et al. 2007). Based on studies in served pathologies and mortalities were due to (1) the other invertebrates, it seems likely that S. marcescens destabilization of the between the coral ani- PDL100 first needs to attach to and then establish itself mal and its photosynthetic dinoflagellates, (2) potential within the coral mucus surface layer. growth promotion of opportunistic pathogens by the The ability to utilize coral mucus as a growth sub- easily metabolizable carbon sources, or (3) a role of strate involves glycosidases, proteases and esterases these sugars in affecting virulence gene regulation in (Vacelet & Thomassin 1991, Krediet et al. 2009). These the existing microbial communities. enzymatic activities are generally consistent with the In order to better understand functions of carbon composition of coral mucus. The major component of sources in the behavior of coral-associated bacteria, the coral mucus surface layer is a sulfated glycopro- the present study tested the hypothesis that simple tein, and mucus also contains a significant amount of sugars, which would be released from the mucus gly- low molecular weight compounds that probably result coprotein by enzymatic digest, affect behaviors in- from microbial degradation of the polymers (Meikle et volved in the interactions of bacteria with mucus of the al. 1988). In mucus of acroporid corals, arabinose, N- coral host (e.g. enzymatic degradation and bacterial acetyl-glucosamine, mannose, glucose, galactose, N- attachment to mucus). acetyl-galactosamine and fucose are the major sugars; The results of the present study demonstrated signif- serine and threonine are the major amino acids; and icant differences in the sequence with which the white Krediet et al.: Catabolite regulation of mucus-degrading enzymes 59

pox pathogen and 2 coral commensals utilized Acrop- assemblies (VivaScience), following manufacturer’s ora palmata mucus in vitro. The white pox pathogen instructions. The low molecular weight fraction was Serratia marcescens PDL100 rapidly downregulates not used in the studies presented here. The high mole- some of its enzymatic activities depending on the cular weight fraction was brought up to volume in arti- availability of simple sugars, and the catabolite repres- ficial seawater (Instant Ocean). sion was relieved within the 2 to 18 h incubation. Feed- Artificial mucus was made by supplementing artifi- back inhibition of enzymatic activities occurred during cial seawater with 169.2 g l–1 L-arabinose, 64.8 g l–1 D- extended incubation (18 h) in the presence of simple mannose, 104.4 g l–1 N-acetyl-glucosamine, 3.6 g l–1 sugars. In the commensal bacteria, catabolite repres- corn starch, 7.2 g l–1 D-galactose (all sugars from Acro¯s sion occurred during the extended incubation (2 to Organics) and 300 g l–1 casamino acids (Sigma 18 h); neither glucose nor N-acetyl-glucosamine had a Aldrich). This mixture crudely approximates the catabolite repressive effect in the coral commensal composition of Acropora mucus reported previously Photobacterium mandapamensis. These differences in (Meikle et al. 1988). the preferred substrates may explain how S. marces- Enzymatic assays. Two overnight cultures of each cens and other opportunistic pathogens become estab- isolate were grown in LB or in GASW broths to an opti- lished within the coral surface mucus layer. cal density at 600 nm (OD600) of 2.0. Cells were pel- leted, washed in filter-sterilized seawater and resus- pended in the same volume of HEPES-buffered MATERIALS AND METHODS seawater. Bacteria were starved in buffered seawater at 30°C while shaking for 3 d. Following starvation, Bacterial strains and culture conditions. Serratia 1 ml of the cell suspension in buffered seawater was marcescens PDL100 is an isolate associated with white added to 2 ml of high molecular weight coral mucus pox of Acropora palmata (Patterson et al. 2002). Photo- preparation with 0.1% (w/v) of one the following sug- bacterium mandapamensis 33C12 and Halomonas ars: D-glucose, N-acetyl-D-glucosamine, D-mannose, L- meridiana 33E7 were isolated from mucus of A. pal- arabinose or D-galactose (Acro¯s Organics). A negative mata based on their ability to grow on a medium sup- control (coral mucus and buffered seawater without plemented with coral mucus (Ritchie 2006). S. mar- supplementation with sugars) was carried out in paral- cescens ATCC43422 (isolated from human throat) was lel. Enzymatic assays were conducted using chro- purchased from the American Type Culture Collection mogenic p-nitro-phenyl substrates (Sigma Aldrich) (ATCC, Manassas, Virginia, USA). after 2 or 18 h of incubation in mucus (at 30°C) follow- Serratia isolates were grown in Luria-Bertani (LB) ing published protocols (Miller 1972). All enzymatic broth (Fisher Scientific). Bacteria isolated from Acrop- reactions were conducted for exactly 24 h; cellular ora palmata mucus were routinely grown in glycerol debris and unused enzymatic substrate were then pel- artificial seawater (GASW) broth (356 mM NaCl, leted at 16 000 × g. The supernatants were then trans-

40 mM MgSO4, 20 mM MgCl2·6H2O, 8 mM KCl, 60 µM ferred to a clear polystyrene 96-well plate and K2HPO4, 33 µM Tris, 7 µM FeSO4, with 0.05% peptone, absorbance at 405 nm (A405) was measured on a Victor- 0.2% yeast extract and 2.0% glycerol; pH 7.0) or on 3 plate reader (Perkin Elmer). Buffered seawater and GASW medium solidified with 1.5% agar (Fisher Sci- coral mucus were included in each plate as blanks. entific) (Ritchie 2006). Unless otherwise stated in text, Enzymatic activities in the coral mucus treatment were seawater used in the experiments was collected at measured and subtracted from all other treatments and Crescent Beach, Florida, USA (29° 45’ 56’’ N, the activity was then calculated using modified Miller

81° 15’ 11’ W) and sterilized through a 0.22 µm filter. As Units (A405/A590). needed for experiments, seawater was buffered with Biofilm assays. For the biofilm assays, bacterial cul-

10 mM HEPES to pH 7. tures were grown overnight to an approximate OD600 Coral mucus was collected from 2 asymptomatic of 2.0. Biofilms were set up either in 3-N-morpholino- Acropora palmata colonies at Looe Key Reef, Florida 2-hydroxpropane sulfonic acid (MOPS) buffered (24° 33’ 75’’ N, 81° 24’ 05’’ W) in August and September colony-forming antigen (CFA) medium or in seawater 2006 with a needleless syringe as in Ritchie (2006). on polystyrene surfaces coated with crude coral Mucus was stored as aliquots at –80°C. As needed, mucus. Buffered CFA medium contained 10 g l–1 –1 –1 aliquots were thawed and exposed to UV irradiation in casamino acids, 1.5 g l yeast extract, 50 mg l MgSO4 –1 plastic tubes (254 nm) for 20 min. Preparations were and 5 mg l MnCl2, buffered with 0.1 M MOPS (pH 7). then filtered through a glass fiber GFC filter, followed To test biofilm formation in CFA, overnight bacterial by filter-sterilization through 0.45 and 0.22 µm filters. cultures were washed and then resuspended in 0.1 M Mucus was then separated into low and high molecu- MOPS-buffered CFA, and 100 µl of the cell suspension lar weight fractions with VisaSpin-15 spin dialysis were added to the wells of a 96-well polystyrene 60 Dis Aquat Org 87: 57–66, 2009

microtiter plate (Fisher Scientific). After 48 h of incuba- During early growth on mucus, mannosidases were not tion, biofilms were stained with 25 µl 1% crystal violet induced in the coral commensals or the human mucosal and biofilm formation was quantified as before (Jack- pathogen S. marcescens ATCC43422, and the constitu- son et al. 2002, Teplitski et al. 2006a). tive levels of mannosidase activities were also low in To test the role of individual mucus monomers in these bacteria (Table 1). Because mannose residues con- attachment to coral mucus-coated surfaces, 50 µl of the nect oligosaccharide decorations to the polypeptide mucus preparation (buffered with 10 mM HEPES to backbone of the mucus glycoprotein (Meikle et al. 1987), pH 7) was incubated in each well of a polystyrene these results may suggest that to utilize coral mucus, the microtiter plate for 1 h, after which 40 µl of the liquid white pox pathogen cleaves off the oligosaccharide side were aspirated and discarded, and the residue was dried at 30°C overnight in Table 1. Enzymatic activities during growth on the high molecular weight a microbiological hood, as in Baving- fraction of mucus from Acropora palmata. Enzymatic activities that increased ton et al. (2004). The inoculum was (italics) or decreased (bold) at least 2-fold after incubation (2 or 18 h) with the prepared by growing bacterial cul- high molecular weight fraction of mucus from A. palmata are indicated tures overnight in LB or in GASW broth at 30°C, then washing them Enzyme Starved 2 h 18 h twice in filter-sterilized buffered sea- Serratia marcescens PDL100 water. One hundred microliters of the N-acetyl-β-D-galactosaminidase 256.21 ± 18.46 171.64 ± 24.00 130.35 ± 5.22 suspension were added to the wells α-D-galactopyranosidase 6.91 ± 0.58 13.14 ± 1.11 1.98 ± 0.64 β coated with mucus. As indicated, -D-galactopyranosidase 13.60 ± 1.07 16.91 ± 1.97 9.11 ± 0.14 α-D-glucopyranosidase 91.70 ± 12.91 60.26 ± 6.04 161.77 ± 1.99 assays were supplemented with 0.1 or β-D-glucopyranosidase 17.03 ± 1.58 16.90 ± 0.44 4.45 ± 0.80 1% (w/v) of individual sugars that α-L-arabinopyranosidase 13.02 ± 1.43 17.86 ± 0.31 11.83 ± 0.08 β occur in coral mucus (D-mannose, L- -L-arabinopyranosidase 4.61 ± 0.33 6.44 ± 0.96 1.86 ± 0.30 α-L-fucopyranosidase 7.70 ± 0.43 10.14 ± 1.18 0.62 ± 0.21 arabinose, D-galactose, N-acetyl-D- β-D-fucopyranosidase 10.81 ± 1.11 17.24 ± 2.65 1.32 ± 0.33 glucosamine). As a control, biofilm for- α-D-mannopyranosidase 3.06 ± 0.70 6.83 ± 1.79 1.72 ± 0.39 mation in wells coated with artificial β-D-mannopyranosidase 9.22 ± 1.71 33.22 ± 5.46 2.89 ± 0.85 mucus was measured in parallel. The Serratia marcescens 43422 β effects of simple sugars on biofilm for- N-acetyl- -D-galactosaminidase 63.99 ± 4.50 146.86 ± 5.76 160.83 ± 6.53 α-D-galactopyranosidase 0.65 ± 0.03 0.22 ± 0.63 0.31 ± 0.53 mation were analyzed through 1-way β-D-galactopyranosidase 4.90 ± 0.20 9.80 ± 0.23 10.31 ± 0.57 ANOVA and Tukey’s Honestly Signif- α-D-glucopyranosidase 40.03 ± 1.92 152.33 ± 6.25 135.59 ± 2.50 β icant Diffference (HSD) for post hoc -D-glucopyranosidase 0.36 ± 0.06 3.68 ± 0.96 3.49 ± 1.00 α-L-arabinopyranosidase 2.48 ± 0.15 11.17 ± 0.02 10.18 ± 0.31 comparison of means. All statistical β-L-arabinopyranosidase 1.16 ± 0.05 0.29 ± 0.26 1.30 ± 0.22 tests were performed using SAS (SAS α-L-fucopyranosidase 0.34 ± 0.03 0.39 ± 0.27 0.50 ± 0.23 Institute). β-D-fucopyranosidase 0.82 ± 0.05 0.55 ± 0.39 0.85 ± 0.34 α-D-mannopyranosidase 0.91 ± 0.04 0.27 ± 0.34 0.97 ± 0.31 β-D-mannopyranosidase 0.88 ± 0.03 1.75 ± 0.82 2.12 ± 0.77 Photobacterium mandapamensis 33C12 RESULTS AND DISCUSSION N-acetyl-β-D-galactosaminidase 8.53 ± 0.24 8.25 ± 0.68 49.08 ± 1.25 α-D-galactopyranosidase 0.85 ± 0.03 0.65 ± 0.52 5.07 ± 0.42 β Enzymatic degradation of the high -D-galactopyranosidase 0.74 ± 0.02 1.03 ± 0.08 2.44 ± 0.01 α-D-glucopyranosidase 1.29 ± 0.06 0.84 ± 0.34 3.88 ± 0.22 molecular weight fraction of coral β-D-glucopyranosidase 0.94 ± 0.07 4.19 ± 0.94 3.69 ± 0.60 mucus α-L-arabinopyranosidase 0.69 ± 0.04 1.19 ± 0.34 2.68 ± 0.28 β-L-arabinopyranosidase 0.99 ± 0.09 1.54 ± 0.27 3.39 ± 0.21 α-L-fucopyranosidase 0.71 ± 0.01 0.56 ± 0.25 2.23 ± 0.26 α D As shown in Table 1, - -glucopyra- β-D-fucopyranosidase 0.74 ± 0.03 1.49 ± 0.39 3.02 ± 0.58 nosidase, N-acetyl-α-D-galactosamini- α-D-mannopyranosidase 0.85 ± 0.25 1.27± 0.39 3.40 ± 0.29 β dase, β-D-galactopyranosidase and α- -D-mannopyranosidase 0.99 ± 0.04 2.42 ± 0.80 2.52 ± 0.82 L-arabinopyranosidase were strongly Halomonas meridiana 33E7 N-acetyl-β-D-galactosaminidase 2.33 ± 0.24 4.52 ± 0.54 6.16 ± 0.30 and constitutively active in Serratia α-D-galactopyranosidase 0.88 ± 0.00 2.13 ± 0.34 2.71 ± 0.62 marcescens PDL100. The correspond- β-D-galactopyranosidase 0.77 ± 0.01 0.11 ± 0.07 0.93 ± 1.04 ing substrates are present within the α-D-glucopyranosidase 313.81 ± 1.67 257.22 ± 13.03 220.89 ± 4.21 β mucus polymer of acroporid corals -D-glucopyranosidase 1.53 ± 0.00 2.05 ± 0.73 1.28 ± 0.44 α-L-arabinopyranosidase 1.19 ± 0.03 0.39 ± 0.31 0.47 ± 0.07 (Meikle et al. 1987). Incubation with the β-L-arabinopyranosidase 1.33 ± 0.00 0.87 ± 0.27 1.00 ± 1.14 high molecular weight fraction of coral α-L-fucopyranosidase 0.91 ± 0.00 0.67 ± 0.33 2.98 ± 0.88 β mucus induced only D-mannopyranosi- -D-fucopyranosidase 0.90 ± 0.04 0.73 ± 0.06 1.18 ± 0.29 α-D-mannopyranosidase 0.87 ± 0.00 1.06 ± 0.26 0.65 ± 0.29 dase activity in this organism as com- β-D-mannopyranosidase 1.14 ± 0.02 1.99 ± 0.81 1.07 ± 0.41 pared to starved cells without mucus. Krediet et al.: Catabolite regulation of mucus-degrading enzymes 61

chains. The cleavage of the oligosaccharide side chains 2008). These results suggested that the pattern and may facilitate enzymatic access to the polypeptide back- temporal regulation of host mucus utilization are cen- bone. Oligosaccharides resulting from this cleavage are tral to the ability of commensals and pathogens to likely taken up by the cells and/or degraded by constitu- establish within the available ecological niches of the tively active glycosidases. host mucus layer. In coral commensal bacteria, incubation with the To begin learning how various components of coral high molecular weight fraction of the Acropora pal- mucus are utilized by commensals and a model mata mucus induced several new activities (Table 1). pathogen, the present study tested the effect of simple The 2 h incubation with the high molecular weight sugars present in coral mucus on the enzymatic activi- fraction of mucus resulted in only modest (2- to 3-fold) ties induced in the white pox pathogen Serratia upregulation of 2 enzymatic activities in Photobac- marcescens PDL100, a human pathogen S. marcescens terium mandapamensis 33C12. Ten new activities ATCC43422 and 2 coral commensals (Photobacterium were induced in this organism during the extended in- mandapamensis 33C12 and Halomonas meridiana cubation on mucus. Only 4 new activities were upreg- 33E7). Glucose, N-acetyl-glucosamine, mannose, ara- ulated in Halomonas meridiana within the same time binose and galactose (at 0.1% w/v) were selected for 3 frame (Table 1). The α-D-glucopyranosidase enzyme reasons: (1) the corresponding residues are known to was constitutively and strongly active in H. meridiana occur in the mucus glycoprotein of acroporid corals (Table 1). This enzyme is induced by the incubation on (Meikle et al. 1987); (2) arabinose, galactose and N- mucus in P. mandapamensis and S. marcescens acetyl-glucosamine are terminal residues in the ATCC43422, suggesting that α-D-glucopyranosidase oligosaccharide decorations of the mucus glycoprotein, may be involved in the degradation of coral mucus. while the mannose residue invariably connects the The differences in the pattern of enzymatic activities oligosaccharide side chain to the polypeptide back- induced during growth of commensals and pathogens bone (Meikle et al. 1987); and (3) glucose is a known on the high molecular weight fraction of coral mucus catabolite repressor of gene regulation in enterobacte- suggest that they rely on different catabolic activities ria and photobacteria (Nealson et al. 1972, Jackson et to establish within their preferred ecological niche. al. 2002, Teplitski et al. 2006b, Deutscher 2008, Gorke The enzymatic hydrolysis of the preferred bonds & Stulke 2008). would release simple sugars, and this in turn may have As shown in Table 2, the addition of simple sugars to a regulatory effect on the enzymatic activities involved mucus-grown cultures of Serratia marcescens PDL100 in mucus degradation. The regulatory effects of vari- had a strong effect during the first 2 h, and the catabo- ous carbon sources on gene expression in bacteria lite repression appears to be relieved — at least par- through catabolite repression are well documented tially — after 18 h of incubation with the simple sugars (Deutscher 2008, Gorke & Stulke 2008). Therefore, we (Table 3). The 2 h incubation on the mucus polymer in tested a potential regulatory role of simple sugars in the presence of glucose strongly repressed activities of the degradation of the mucus polymer. 8 (out of 11 tested) enzymatic activities by 2- to 144- fold in S. marcescens PDL100. Activities of fucopyra- nosidase and mannopyranosidase were repressed the Catabolite control of coral mucus degradation by strongest. The addition of N-acetyl-glucosamine or simple sugars mannose had similarly strong repressive effects on enzymatic activities in S. marcescens PDL100. Treat- Commensal and pathogenic bacteria utilize compo- ment with arabinose or galactose had a lesser effect on nents of host mucus using different strategies, and the enzymatic activities in S. marcescens PDL100 these differences may help them occupy different eco- (Table 2) and the treatment with glucose, arabinose or logical niches. For example, commensal Escherichia galactose strongly upregulated α-D-glucopyranosidase coli and enterohaemorrhagic E. coli O157:H7 are capa- during the first 2 h. These results suggest that arabi- ble of utilizing essentially the same carbon sources nose and galactose are not the preferred carbon (Chang et al. 2004, Fabich et al. 2008). However, tem- sources. N-acetyl-galactosaminidase and α-D-glucopy- poral regulation of enzymatic activities and timing of ranosidase were not subject to catabolite repression carbon source utilization during growth on the compo- (Table 2). These results suggest a mechanism by which nents of mouse intestinal mucus were different in these S. marcescens PDL100 may outcompete the native 2 conspecifics. Mutational inactivation of the corre- microbiota: during the starvation period it strongly sponding pathways affected fitness of the bacteria dur- upregulates several enzymatic activities and — when ing colonization of the mouse gut. The same mutations the substrates are present — these enzymes hydrolyze in the commensal and enterohaemorrhagic E. coli had the corresponding substrates. The availability of pre- different effects on fitness of the 2 strains (Fabich et al. ferred carbon sources downregulates other enzymatic 62 Dis Aquat Org 87: 57–66, 2009

Table 2. Enzymatic activities detected during the 2 h incubation on the high molecular weight fraction of mucus (M) from Acropora palmata in the presence of simple sugars. Simple sugars added: galactose (Gal), glucose (Glu), arabinose (Ara), mannose (Man) and N-acetyl-glucosamine (NGlu). Enzymatic activities that increased (italics) or decreased (bold) at least 2-fold after incubation are indicated

Enzyme M M + Gal M + Glu M + Ara M + Man M + NGlu

Serratia marcescens PDL100 N-acetyl-β-D-galactosaminidase 171.64 ± 24.00 176.45 ± 5.36 158.72 ± 5.22 177.44 ± 0.29 97.45 ± 3.76 118.42 ± 5.44 α-D-galactopyranosidase 13.14 ± 1.11 1.44 ± 0.62 0.74 ± 0.64 0.90 ± 0.64 0.19 ± 0.61 0.85 ± 0.41 β-D-galactopyranosidase 16.91 ± 1.97 20.96 ± 0.46 8.29 ± 0.04 24.96 ± 0.30 2.39 ± 0.12 3.27 ± 0.33 α-D-glucopyranosidase 60.26 ± 6.04 163.10 ± 5.02 244.83 ± 2.01 166.55 ± 2.75 36.69 ± 0.12 40.52 ± 15.81 β-D-glucopyranosidase 16.90 ± 0.44 4.07 ± 0.95 3.33 ± 1.01 4.05 ± 1.02 2.91 ± 0.89 8.03 ± 0.20 α-L-arabinopyranosidase 17.86 ± 0.31 17.36 ± 0.11 11.46 ± 0.20 18.85 ± 0.18 0.90 ± 0.29 1.73 ± 0.11 β-L-arabinopyranosidase 6.44 ± 0.96 1.58 ± 0.28 1.22 ± 0.29 1.63 ± 0.32 0.33 ± 0.24 0.08 ± 0.13 α-L-fucopyranosidase 10.14 ± 1.18 0.76 ± 0.27 0.07 ± 0.29 0.81 ± 0.29 2.88 ± 0.19 4.32 ± 0.29 β-D-fucopyranosidase 17.24 ± 2.65 0.90 ± 0.36 0.78 ± 0.40 0.92 ± 0.41 0.54 ± 0.38 0.07 ± 0.24 α-D-mannopyranosidase 6.83 ± 1.79 1.49 ± 0.38 1.01 ± 0.38 1.55 ± 0.41 0.97 ± 0.40 0.41 ± 0.15 β-D-mannopyranosidase 33.22 ± 5.46 2.64 ± 0.83 2.03 ± 0.82 2.70 ± 0.87 2.04 ± 0.84 1.57 ± 0.70 Serratia marcescens 43422 N-acetyl-β-D-galactosaminidase 146.86 ± 5.76 165.29 ± 1.23 106.16 ± 1.78 168.07 ± 0.20 191.86 ± 6.18 110.79 ± 4.38 α-D-galactopyranosidase 0.22 ± 0.63 1.58 ± 0.62 0.31 ± 0.59 0.88 ± 0.61 0.72 ± 0.61 0.05 ± 0.12 β-D-galactopyranosidase 9.80 ± 0.23 9.54 ± 0.23 7.19 ± 0.20 8.16 ± 0.01 4.15 ± 0.05 7.14 ± 0.15 α-D-glucopyranosidase 152.33 ± 6.25 82.21 ± 4.36 114.81 ± 6.45 82.93 ± 0.97 91.39 ± 2.91 130.16 ± 2.61 β-D-glucopyranosidase 3.68 ± 0.96 4.39 ± 1.00 3.93 ± 0.95 4.37 ± 1.01 4.21 ± 1.02 4.03 ± 0.95 α-L-arabinopyranosidase 11.17 ± 0.02 8.63 ± 0.17 7.58 ± 0.07 7.32 ± 0.80 3.76 ± 0.22 7.47 ± 0.05 β-L-arabinopyranosidase 0.29 ± 0.26 1.50 ± 0.32 0.50 ± 0.25 1.42 ± 0.31 1.34 ± 0.28 0.48 ± 0.03 α-L-fucopyranosidase 0.39 ± 0.27 0.80 ± 0.29 0.39 ± 0.27 0.75 ± 0.28 0.71 ± 0.28 0.64 ± 0.26 β-D-fucopyranosidase 0.55 ± 0.39 1.16 ± 0.32 0.77 ± 0.41 0.14 ± 0.63 1.26 ± 0.42 0.99 ± 0.37 α-D-mannopyranosidase 0.27 ± 0.34 1.24 ± 0.40 0.45 ± 0.40 1.16 ± 0.37 1.05 ± 0.34 0.65 ± 0.28 β-D-mannopyranosidase 1.75 ± 0.82 2.56 ± 0.85 1.88 ± 0.85 2.45 ± 0.84 2.38 ± 0.84 1.53 ± 0.45 Photobacterium mandapamensis 33C12 N-acetyl-β-D-galactosaminidase 8.25 ± 0.68 51.75 ± 0.70 12.36 ± 0.37 50.94 ± 0.44 51.33 ± 0.38 11.00 ± 0.57 α-D-galactopyranosidase 0.65 ± 0.52 0.06 ± 0.62 0.37 ± 0.53 2.36 ± 0.29 2.23 ± 0.36 0.66 ± 0.59 β-D-galactopyranosidase 1.03 ± 0.08 0.99 ± 0.13 0.58 ± 0.01 0.67 ± 0.29 0.93 ± 0.10 1.09 ± 0.11 α-D-glucopyranosidase 0.84 ± 0.34 0.64 ± 0.38 0.64 ± 0.29 0.52 ± 0.35 0.11 ± 0.11 0.90 ± 0.40 β-D-glucopyranosidase 4.19 ± 0.94 2.47 ± 0.06 4.04 ± 0.97 3.45 ± 1.00 3.41 ± 0.99 4.23 ± 0.99 α-L-arabinopyranosidase 1.19 ± 0.34 0.96 ± 0.33 1.03 ± 0.36 0.93 ± 0.37 1.01 ± 0.34 1.24 ± 0.37 β-L-arabinopyranosidase 1.54 ± 0.27 1.55 ± 0.32 1.39 ± 0.27 1.43 ± 0.29 1.46 ± 0.31 1.62 ± 0.31 α-L-fucopyranosidase 0.56 ± 0.25 0.26 ± 0.18 0.42 ± 0.27 0.29 ± 0.28 0.23 ± 0.21 0.67 ± 0.28 β-D-fucopyranosidase 1.49 ± 0.39 1.47 ± 0.44 0.62 ± 1.29 1.31 ± 0.41 1.33 ± 0.39 1.57 ± 0.42 α-D-mannopyranosidase 1.27 ± 0.34 1.38 ± 0.84 1.09 ± 0.36 1.24 ± 0.38 1.34 ± 0.39 1.35 ± 0.40 β-D-mannopyranosidase 2.42 ± 0.80 2.21 ± 0.84 2.21 ± 0.82 1.96 ± 0.78 2.18 ± 0.83 2.46 ± 0.85 Halomonas meridiana 33E7 N-acetyl-β-D-galactosaminidase 4.52 ± 0.54 1.20 ± 2.43 2.53 ± 0.08 3.38 ± 1.02 2.66 ± 1.20 1.73 ± 1.20 α-D-galactopyranosidase 2.13 ± 0.34 0.51 ± 0.20 1.14 ± 0.33 0.71 ± 0.49 0.46 ± 0.60 2.21 ± 0.25 β-D-galactopyranosidase 0.11 ± 0.07 0.60 ± 0.34 0.87 ± 0.07 1.26 ± 0.07 1.05 ± 0.07 1.88 ± 2.16 α-D-glucopyranosidase 257.22 ± 13.03 164.27 ± 10.66 177.93 ± 21.04 148.72 ± 12.67 161.84 ± 8.56 213.47 ± 3.68 β-D-glucopyranosidase 2.05 ± 0.73 2.67 ± 0.63 3.18 ± 0.81 3.45 ± 0.76 2.45 ± 0.87 1.66 ± 0.18 α-L-arabinopyranosidase 0.39 ± 0.31 0.25 ± 0.03 0.63 ± 0.29 1.02 ± 0.29 0.04 ± 0.34 0.70 ± 0.34 β-L-arabinopyranosidase 0.87 ± 0.27 1.15 ± 0.17 1.15 ± 0.21 1.53 ± 0.22 1.37 ± 0.29 1.40 ± 0.30 α-L-fucopyranosidase 0.67 ± 0.33 0.31 ± 0.14 0.02 ± 0.05 0.45 ± 0.15 0.46 ± 0.26 0.39 ± 0.24 β-D-fucopyranosidase 0.73 ± 0.06 1.41 ± 0.29 1.48 ± 0.24 1.64 ± 0.34 1.43 ± 0.30 1.39 ± 0.27 α-D-mannopyranosidase 1.06 ± 0.26 1.40 ± 0.34 0.96 ± 0.26 1.61 ± 0.34 1.31 ± 0.29 1.38 ± 0.37 β-D-mannopyranosidase 1.99 ± 0.81 2.14 ± 0.71 2.12 ± 0.80 2.58 ± 0.76 1.95 ± 0.45 2.37 ± 0.84

activities. This hypothesis is yet to be tested in situ lize coral mucus. Based on the results presented in using defined mutants. Table 2, it appears that the 2 h incubation of the 2 com- The 2 coral commensals tested in the present study mensals on mucus with simple sugars did not have a appear to rely on mucus degradation strategies that strong catabolite repression effect on the enzymatic are distinct from the white pox pathogen, although this activities. Treatment of Photobacterium mandapamen- limited survey of coral commensals does not suggest a sis with mannose, arabinose or galactose stimulated N- unifying model by which commensals as a group uti- acetyl galactosaminidase, and after 18 h of incubation Krediet et al.: Catabolite regulation of mucus-degrading enzymes 63

Table 3. Enzymatic activities during the 18 h incubation on the high molecular weight fraction of mucus (M) from Acropora palmata in the presence of simple sugars. Simple sugars added: galactose (Gal), glucose (Glu), arabinose (Ara), mannose (Man) and N-acetyl-glucosamine (NGlu). Enzymatic activities that increased (italics) or decreased (bold) at least 2-fold after incubation are indicated

Enzyme M M + Gal M + Glu M + Ara M + Man M + NGlu

Serratia marcescens PDL100 N-acetyl-β-D-galactosaminidase 130.35 ± 3.87 108.6 ± 4.72 38.16 ± 5.64 192.98 ± 5.95 111.31 ± 0.73 82.43 ± 0.12 α-D-galactopyranosidase 1.98 ± 0.64 1.79 ± 0.63 1.31 ± 0.63 0.94 ± 0.65 1.31 ± 0.63 1.69 ± 0.64 β-D-galactopyranosidase 9.11 ± 0.14 45.61 ± 0.01 2.11 ± 0.03 14.42 ± 0.15 5.85 ± 0.07 2.75 ± 0.08 α-D-glucopyranosidase 161.77 ± 1.99 13.84 ± 0.44 76.16 ± 3.77 15.60 ± 1.48 9.61 ± 0.10 7.28 ± 0.18 β-D-glucopyranosidase 4.45 ± 0.80 4.79 ± 0.97 2.69 ± 0.94 4.24 ± 0.98 4.63 ± 0.99 4.93 ± 1.01 α-L-arabinopyranosidase 11.83 ± 0.08 40.88 ± 0.01 2.40 ± 0.28 15.62 ± 0.07 6.07 ± 0.31 3.05 ± 0.28 β-L-arabinopyranosidase 1.86 ± 0.30 1.27 ± 0.26 1.18 ± 0.28 0.92 ± 0.30 1.44 ± 0.29 1.69 ± 0.29 α-L-fucopyranosidase 0.62 ± 0.21 1.02 ± 0.29 0.62 ± 0.08 0.82 ± 0.29 0.89 ± 0.25 1.02 ± 0.30 β-D-fucopyranosidase 1.32 ± 0.33 0.57 ± 0.41 1.25 ± 0.33 0.94 ± 0.40 1.40 ± 0.37 1.77 ± 0.42 α-D-mannopyranosidase 1.72 ± 0.39 1.59 ± 0.40 1.62 ± 0.38 1.21 ± 0.40 1.43 ± 0.37 1.70 ± 0.40 β-D-mannopyranosidase 2.89 ± 0.85 2.92 ± 0.85 2.81 ± 0.86 2.48 ± 0.86 2.80 ± 0.85 2.99 ± 0.83 Serratia marcescens 43422 N-acetyl-β-D-galactosaminidase 160.83 ± 6.53 181.11 ± 10.59 88.83 ± 3.88 174.67 ± 6.27 147.03 ± 5.11 68.07 ± 0.08 α-D-galactopyranosidase 0.31 ± 0.53 1.48 ± 0.58 0.65 ± 0.54 0.82 ± 0.57 1.15 ± 0.62 1.46 ± 0.55 β-D-galactopyranosidase 10.31 ± 0.57 10.62 ± 0.87 0.67 ± 0.03 8.54 ± 0.32 2.76 ± 0.02 3.76 ± 0.20 α-D-glucopyranosidase 135.59 ± 2.50 90.47 ± 9.16 0.33 ± 0.27 86.16 ± 3.69 70.33 ± 2.45 16.62 ± 1.79 β-D-glucopyranosidase 3.49 ± 1.00 4.25 ± 0.92 4.02 ± 0.84 4.31 ± 0.96 4.59 ± 1.02 4.69 ± 0.85 α-L-arabinopyranosidase 10.18 ± 0.31 9.62 ± 0.41 0.43 ± 0.30 7.59 ± 0.76 2.45 ± 0.24 3.78 ± 0.02 β-L-arabinopyranosidase 1.30 ± 0.22 1.39 ± 0.26 1.57 ± 0.28 1.37 ± 0.27 1.65 ± 0.29 2.22 ± 0.31 α-L-fucopyranosidase 0.50 ± 0.23 0.74 ± 0.25 0.63 ± 0.26 0.72 ± 0.25 0.89 ± 0.28 1.17 ± 0.29 β-D-fucopyranosidase 0.85 ± 0.34 1.02 ± 0.24 1.43 ± 0.40 0.01 ± 0.76 1.55 ± 0.42 1.90 ± 0.42 α-D-mannopyranosidase 0.97 ± 0.31 1.12 ± 0.33 1.12 ± 0.36 1.11 ± 0.33 1.36 ± 0.36 1.96 ± 0.41 β-D-mannopyranosidase 2.12 ± 0.77 2.45 ± 0.79 2.36 ± 0.80 2.39 ± 0.79 2.70 ± 0.85 3.27 ± 0.86 Photobacterium mandapamensis 33C12 N-acetyl-β-D-galactosaminidase 49.08 ± 1.25 66.60 ± 3.21 54.16 ± 4.07 53.67 ± 6.77 59.01 ± 2.44 68.98 ± 1.59 α-D-galactopyranosidase 5.07 ± 0.42 2.37 ± 0.10 8.07 ± 1.07 3.92 ± 0.68 2.07 ± 0.24 18.58 ± 0.94 β-D-galactopyranosidase 2.44 ± 0.01 1.17 ± 1.53 2.74 ± 0.24 0.36 ± 0.00 0.74 ± 0.06 2.57 ± 0.18 α-D-glucopyranosidase 3.88 ± 0.22 0.38 ± 0.01 5.16 ± 0.41 0.67 ± 0.39 0.04 ± 0.25 3.82 ± 0.10 β-D-glucopyranosidase 3.69 ± 0.60 1.28 ± 0.09 2.99 ± 0.17 2.95 ± 0.60 3.03 ± 0.84 3.42 ± 0.42 α-L-arabinopyranosidase 2.68 ± 0.28 0.59 ± 0.27 8.37 ± 2.26 0.77 ± 0.26 0.71 ± 0.23 3.45 ± 0.83 β-L-arabinopyranosidase 3.39 ± 0.21 0.98 ± 0.22 2.43 ± 0.26 0.50 ± 1.26 1.23 ± 0.17 2.29 ± 0.03 α-L-fucopyranosidase 2.23 ± 0.26 1.49 ± 1.01 2.41 ± 0.05 0.30 ± 0.16 0.02 ± 0.09 4.28 ± 0.48 β-D-fucopyranosidase 3.02 ± 0.58 0.87 ± 1.06 2.09 ± 0.06 0.94 ± 0.10 0.41 ± 0.17 5.44 ± 2.00 α-D-mannopyranosidase 3.40 ± 0.29 0.62 ± 0.29 2.75 ± 0.28 4.82 ± 3.31 1.18 ± 0.32 2.14 ± 0.12 β-D-mannopyranosidase 2.52 ± 0.82 1.52 ± 0.60 1.86 ± 0.62 1.33 ± 0.26 1.27 ± 0.34 3.63 ± 0.68 Halomonas meridiana 33E7 N-acetyl-β-D-galactosaminidase 6.16 ± 0.30 2.57 ± 0.79 1.05 ± 0.96 3.10 ± 1.00 3.65 ± 0.97 3.40 ± 1.18 α-D-galactopyranosidase 2.71 ± 0.62 0.56 ± 0.28 0.27 ± 0.60 1.72 ± 0.41 1.00 ± 0.58 1.18 ± 0.28 β-D-galactopyranosidase 0.93 ± 1.04 1.01 ± 1.00 0.70 ± 0.06 0.06 ± 0.63 1.28 ± 0.10 0.58 ± 0.02 α-D-glucopyranosidase 220.89 ± 4.21 93.78 ± 15.78 100.47 ± 21.58 124.70 ± 4.31 110.93 ± 12.40 188.85 ± 30.67 β-D-glucopyranosidase 1.18 ± 0.44 2.15 ± 0.00 3.09 ± 0.84 3.78 ± 0.99 3.76 ± 0.98 2.58 ± 0.92 α-L-arabinopyranosidase 0.47 ± 0.07 0.23 ± 0.05 0.05 ± 0.03 0.88 ± 0.27 0.29 ± 0.81 0.25 ± 0.32 β-L-arabinopyranosidase 1.00 ± 1.14 1.24 ± 0.17 2.52 ± 0.187 1.59 ± 0.20 1.58 ± 0.12 0.96 ± 0.09 α-L-fucopyranosidase 2.98 ± 0.88 0.05 ± 0.26 0.28 ± 0.12 0.42 ± 0.08 0.78 ± 0.27 0.53 ± 0.07 β-D-fucopyranosidase 1.18 ± 0.29 1.11 ± 0.06 1.61 ± 0.37 1.62 ± 0.36 1.85 ± 0.43 1.39 ± 0.41 α-D-mannopyranosidase 0.65 ± 0.29 1.25 ± 0.27 1.27 ± 0.27 1.65 ± 0.36 1.62 ± 0.26 0.66 ± 0.01 β-D-mannopyranosidase 1.07 ± 0.41 2.48 ± 0.70 1.66 ± 0.35 2.74 ± 0.84 2.73 ± 0.84 1.66 ± 0.67

the induction was relieved (Table 3). Neither glucose (Table 2), suggesting that the corresponding substrates nor N-acetyl-glucosamine had a strong catabolite maybe preferentially co-utilized by the bacterium. repression effect on P. mandapamensis (Tables 2 & 3). Within the 18 h incubation, the catabolite repression This suggests that neither glucose nor N-acetyl-glu- effects of galactose, arabinose and mannose on the cosamine is the preferred carbon source for this bac- enzymatic activities in this bacterium were the terium. Both galactosidase and N-acetyl-galactosidase strongest; galactosidases, arabinosidases and α-D- appear to be co-upregulated within the first 2 h mannopyranosidase appear to be controlled by the 64 Dis Aquat Org 87: 57–66, 2009

feedback inhibition in the presence of sugars that within coral mucus are important to the outcome of could result from the activity of these enzymes coral–bacteria interactions. Because bacteria utilize (Table 3). Besides N-acetyl-glucosamine, mannose, mucus as a growth substrate, we tested bacterial arabinose and galactose are the 3 most abundant sug- biofilm formation on mucus in the presence of mono- ars in the mucus polymer of acroporid corals (Meikle et saccharides that would result from the enzymatic al. 1987). Arabinose and galactose occur as terminal hydrolysis of mucus polymers. residues, while mannose links the oligosaccharide side Biofilm assays were set up in polystyrene microtiter chain to the polypeptide backbone (Meikle et al. 1987). plates coated with either crude coral mucus or with These sugars are most likely released during the enzy- artificial mucus (a solution containing carbon and matic hydrolysis of the polymer through the actions of nitrogen sources to approximate the composition and endo- and exo-glycanases in situ. The strong, but viscosity of coral mucus). The addition of sugars at delayed, repressive effect of these sugars may, in part, 0.1% had no effect on biofilm formation (data not explain why growth of P. mandapamensis 33C12 on shown). As shown in Fig. 1A, Serratia marcescens coral mucus plateaus early despite the fact that it ini- PDL100 formed biofilms on artificial mucus and on tially grows fast on coral mucus (Krediet et al. 2009). crude Acropora palmata mucus, regardless of the addi- The abundance of simple sugars resulting from hydro- tion of simple sugars to the sample. Biofilm formation lysis of the mucus polymer may signal to this commen- by a human pathogen S. marcescens 43422 was sal that its habitat (coral mucus surface layer) is stronger on artificial mucus than on any combination of degraded, and the bacterium switches to a lower meta- coral mucus (Fig. 1B). In control experiments, crude bolic rate. Most pathogens, on the other hand, in mucus of A. palmata also modestly inhibited biofilm response to nutrient limitation, upregulate virulence- formation in CFA medium (data not shown), suggest- related genes and thus may switch from saprophytic ing that an unknown component of coral mucus growth to virulence (Teplitski et al. 2006b, Gorke & inhibits biofilm formation in some bacteria. This is Stulke 2008). reminiscent of a study in which mucus of some echino- In another coral commensal, Halomonas meridiana derms contained substances capable of inhibiting 33E7, galactose and glucose had some catabolite re- attachment by bacteria (Bavington et al. 2004). pression effect after 2 h of incubation on mucus, In Photobacterium mandapamensis there was no dif- although the number of repressed activities and the ference in biofilm formation on mucus-coated surfaces magnitude of repression were less than in Serratia (vs. surfaces coated with artificial mucus). As shown in marcescens PDL100 (Table 2). With the exception of N- Fig. 1D, biofilm formation on mucus-coated surfaces acetyl-glucosamine, the addition of all monosaccha- and those formed on mucus-coated surfaces in the rides promoted β-galactosidase activity (Table 2) after presence of simple sugars was similar (although statis- 2 h. These results suggest that galactose and glucose tically significant, 2-fold differences were observed may be preferred carbon sources based on their strong between biofilms formed on mucus-coated surfaces catabolite repression. with the addition of 1% arabinose or galactose, com- The 4 tested bacteria had a different set of preferred pared to the biofilms formed in the presence of glu- carbon sources when grown on the mucus polymers; cose, N-acetyl glucosamine or mannose). Biofilm for- they also differed in the strength of their enzymatic mation by Halomonas meridiana 33E7 on surfaces activities. These differences may help explain how coated with mucus was somewhat lower compared to coral mucus-associated microbial communities are the biofilm formation on surfaces coated with the arti- structured in situ, and how pathogens may be able to ficial mucus (Fig. 1C). With the exception of glucose, establish with native coral microbiota. Further muta- none of the simple sugars had a statistically significant tional analyses are needed to establish the role for effect on biofilm formation. Interestingly, glucopyra- these activities in situ. nosidase (an enzyme that makes glucose available to the cell) is the strongest constitutively active enzyme in H. meridiana 33E7. It is, therefore, possible that the Attachment to coral mucus availability of the preferred carbon source (glucose for H. meridiana, galactose or arabinose for Photobac- Adhesion of a coral pathogen Vibrio shiloi to mucus terium mandapamensis, based on the strongest of its coral host was shown to be one of the steps catabolite repression activity by these sugars) indicates involved in pathogenesis (Banin et al. 2001, Rosenberg a preferred ecological niche for this organism and & Falkovitz 2004). In the reef environment, coral stimulates settlement and biofilm formation. An alter- mucus is colonized by microbes and microscopic euka- native explanation for the observed effect could be ryotes (Vacelet & Thomassin 1991). These observations based on the reports that bacteria bind to specific car- suggest that binding to mucus and establishment bohydrate receptors present in coral mucus (Banin et Krediet et al.: Catabolite regulation of mucus-degrading enzymes 65

Fig. 1. Biofilm formation on surfaces coated with mucus of Acropora palmata in the presence of simple sugars. (A) White pox pathogen Serratia marcescens PDL100, (B) human pathogen S. marcescens ATCC43422, coral commensals (C) Halomonas meridiana 33E7 and (D) Photobacterium mandapamensis 33E12. A590: absorbance at 590 nm. Man: D-mannose; Ara: L-arabinose; Gal: D-galactose; NGlu: N-acetyl-D-glucosamine; Glu: glucose. Error bars represent SE; data points that are statistically different are indicated by different lowercase letters al. 2001) and the addition of simple sugars may either sion was relieved within 18 h of incubation on mucus, facilitate or inhibit these interactions. suggesting that as nutrients become less available, this pathogen becomes more aggressive. This would be consistent with the observations in other pathogens CONCLUSIONS that upregulate their virulence genes when preferred carbon sources become limiting (Deutscher 2008, It is becoming increasingly clear that nutrients and Gorke & Stulke 2008, Teplitski et al. 2006b). signals found in the coral surface mucopolysaccharide A comparison of enzymatic activities induced in the layer dictate, at least in part, the structure and the com- commensals during growth on the mucus polymer sug- position of the associated microbiota (Ritchie 2006, gested that they relied on different strategies to colo- Sharon & Rosenberg 2008, Teplitski & Ritchie 2009). nize the coral surface mucopolysaccharide layer. In The present study suggests that to colonize mucus of its Photobacterium mandapamensis at least 10 new gly- coral host, Acropora palmata, a model pathogen Serra- cosidases were induced at various stages of in vitro tia marcescens and 2 representative coral commensals growth on the mucus polymers. The availability of relied on different strategies. Upon exposure to coral galactose, mannose or arabinose had a strong re- mucus, mannopyranosidases were induced in S. pressive affect on enzymatic activities after 18 h of marcescens, these enzymes hydrolyze oligosaccharide growth on mucus. In another commensal bacterium, side chains from the polypeptide backbone of the mu- Halomonas meridiana, α-D-glucopyranosidase was cus glycoprotein. At least 5 glycosidases were strongly strongly and constitutively produced and 4 new gly- expressed in the starved cultures of this bacterium, and cosidases were induced during growth on the high these enzymes may help degrade mucus oligosaccha- molecular weight fraction of coral mucus. Interestingly, rides that are released by mannosidases. Most of the in both commensals, catabolite repression effects were glycosidases in S. marcescens were subject to catabo- strongest after 18 h of incubation, while in the white lite repression early, and this suggests that the ability to pox pathogen, the catabolite repression effects were efficiently downregulate enzymatic activities during largely relieved within the same time frame. It may be colonization of coral mucus may help this pathogen es- possible (while not yet experimentally tested) that this tablish within the coral mucus layer. Catabolite repres- catabolite repression by sugars may arrest overgrowth 66 Dis Aquat Org 87: 57–66, 2009

of commensals, which could be detrimental to the coral and by environmental and coral commensal bacteria. Appl host. Elucidation of the differences in the strategies Environ Microbiol 75(12):3851-3858 that coral commensals and opportunistic pathogens Kuntz NM, Kline DI, Sandin SA, Rohwer F (2005) Pathologies and mortality rates caused by organic carbon and nutrient will help define the mechanisms of coral disease and stressors in three Caribbean coral species. Mar Ecol Prog may lead to research on defining potential approaches Ser 294:173–180 for managing or treating coral diseases. Kurz CL, Chauvet S, Andres E, Aurouze M and others (2003) Virulence factors of the human opportunistic pathogen Serratia marcescens identified by in vivo screening. Acknowledgements. This research, under Current Research EMBO J 22:1451–1460 Information System project FLA-SWS-04591 of the Florida Meikle P, Richards GN, Yellowlees D (1987) Structural deter- Agricultural Experiment Station, was supported by Florida mination of the oligosaccharide side chains from a glyco- Sea Grant #NA060AR4170014 (to M.T. and K.B.R.), funding protein isolated from the mucus of the coral Acropora for- from Florida specialty license plate program Protect Our mosa. J Biol Chem 262:16941–16947 Reefs (to K.B.R., M.T. and C.J.K.), the Charles A. and Anne Meikle P, Richards GN, Yellowlees D (1988) Structural inves- Morrow Lindbergh Foundation award (to K.B.R. and M.T.), tigations on the mucus from six species of coral. Mar Biol National Geographic Society grant #8207-07 and a Seed 99:187–193 Grant from the University of Florida School of Natural Miller JH (1972) Experiments in molecular genetics. Cold Resources and Environment (UF SNRE). C.J.K. is a UF SNRE Spring Harbor Laboratory, Cold Spring Harbor, NY Alumni Fellow. Molchanova VI, Ovodova RG, Ovodov YS, Elkin YN, Fernan- dezsantana V (1985) Studies of the polysaccharide moiety of corallan, a glycoprotein from Pseudopterogorgia ameri- LITERATURE CITED cana. Carbohydr Res 141:289–293 Nealson KH, Eberhard A, Hastings JW (1972) Catabolite Banin E, Israely T, Fine M, Loya Y, Rosenberg E (2001) Role of repression of bacterial bioluminescence: functional impli- endosymbiotic zooxanthellae and coral mucus in the cations. Proc Natl Acad Sci USA 69:1073–1076 adhesion of the coral-bleaching pathogen Vibrio shiloi to Nehme NT, Liegeois S, Kele B, Giammarinaro P and others its host. FEMS Microbiol Lett 199:33–37 (2007) A model of bacterial intestinal infections in Bavington CD, Lever R, Mulloy B, Grundy MM, Page CP, Drosophila melanogaster. PLoS Pathog 3:e173 Richardson NV, McKenzie JD (2004) Anti-adhesive glyco- Patterson KL, Porter JW, Ritchie KB, Polson SW and others proteins in echinoderm mucus secretions. Comp Biochem (2002) The etiology of white pox, a lethal disease of the Physiol B 139:607–617 Caribbean elkhorn coral, Acropora palmata. Proc Natl Chang DE, Smalley DJ, Tucker DL, Leatham MP and others Acad Sci USA 99:8725–8730 (2004) Carbon nutrition of Escherichia coli in the mouse Ritchie KB (2006) Regulation of microbial populations by coral intestine. Proc Natl Acad Sci USA 101:7427–7432 surface mucus and mucus-associated bacteria. Mar Ecol Deutscher J (2008) The mechanisms of carbon catabolite Prog Ser 322:1–14 repression in bacteria. Curr Opin Microbiol 11:87–93 Ritchie KB, Smith GW (2004) Microbial communities of coral Ducklow HW, Mitchell R (1979) Composition of mucus surface mucopolysaccharide layers. In: Rosenberg E, Loya released by coral reef coelenterates. Limnol Oceanogr 21: Y (eds) Coral health and diseases. Springer-Verlag, New 706–714 York, p 259–264 Fabich AJ, Jones SA, Chowdhury FZ, Cernosek A and others Rosenberg E, Falkovitz L (2004) The Vibrio shiloi/Oculina (2008) Comparison of carbon nutrition for pathogenic and patagonica model system of coral bleaching. Annu Rev commensal Escherichia coli strains in the mouse intestine. Microbiol 58:143–159 Infect Immun 76:1143–1152 Sharon G, Rosenberg E (2008) Bacterial growth on coral Gorke B, Stulke J (2008) Carbon catabolite repression in bac- mucus. Curr Microbiol 56:481–488 teria: many ways to make the most out of nutrients. Nat Teplitski M, Ritchie KB (2009) How feasible is the biological Rev Microbiol 6:613–624 control of coral diseases? Trends Ecol Evol 24:378–385 Jackson DW, Simecka JW, Romeo T (2002) Catabolite repres- Teplitski M, Al-Agely A, Ahmer BM (2006a) Contribution of sion of Escherichia coli biofilm formation. J Bacteriol the SirA regulon to biofilm formation in Salmonella enter- 184:3406–3410 ica serovar Typhimurium. Microbiology 152:3411–3424 Klaus JS, Janse I, Heikoop JM, Sanford RA, Fouke BW (2007) Teplitski M, Goodier RI, Ahmer BM (2006b) Catabolite Coral microbial communities, zooxanthellae and mucus repression of the SirA regulatory cascade in Salmonella along gradients of seawater depth and coastal pollution. enterica. Int J Med Microbiol 296:449–466 Environ Microbiol 9:1291–1305 Vacelet E, Thomassin BA (1991) Microbial utilization of coral Koren O, Rosenberg E (2006) Bacteria associated with mucus mucus in long-term in situ incubation over a coral reef. and tissues of the coral Oculina patagonica in summer and Hydrobiologia 211:19–32 winter. Appl Environ Microbiol 72:5254–5259 Vine NG, Leukes WD, Kaiser H (2004) In vitro growth charac- Krediet CJ, Ritchie KB, Cohen M, Lipp EK, Sutherland KP, teristics of five candidate aquaculture probiotics and two Teplitski M (2009) Utilization of mucus from the coral fish pathogens grown in fish intestinal mucus. FEMS Acropora palmata by the pathogen Serratia marcescens Microbiol Lett 231:145–152

Editorial responsibility: Kiho Kim, Submitted: January 6, 2009; Accepted: April 28, 2009 Washington, DC, USA Proofs received from author(s): June 22, 2009